The purpose of this project is to document and clarify reproducibility issues arising in selected computational results, with particular attention to Figure 1 and parts of Figure 2. We provide a reproducibility note, MATLAB code, and supporting documentation to make the computational analysis transparent and verifiable.
Our replication analysis indicates that the plotted surface appears to be transposed relative to the axis labels. In MATLAB, contourf(X,Y,Z) expects rows of Z to correspond to Y and columns of Z to correspond to X. In the relevant scripts (including bigdatagrowsigma.m and bigdatagrowxi.m), the matrix construction and plotting conventions do not appear to align, which affects the interpretation of the plotted contours.
Transitional dynamics: In our computational analysis of the central planner economy, we do not find convergence of the R&D labor share to a unique steady state. Instead, the system appears to admit a continuum of steady states. Under this interpretation, economies may converge in long-run growth rates while differing in long-run labor allocations to research and in aggregate levels. Our reproducibility note discusses the implications of this feature for interpreting the reported transitional dynamics. For more details, see our paper here .
We provide MATLAB scripts and documentation for the transitional dynamics analysis reported in our reproducibility study here: the script Part_1.m produces Figure 1 and 3 in our paper, whereas script Part_2.m produces Figure 2 in our paper. For all computations, we have used MATLAB 2024a. Python codes will follow. Please feel free to reach out to me if you have any questions.